AI Feels Like the Pandemic All Over Again

For some reason, this current era of AI development recently reminds me of the first years of the Covid pandemic. Not in the way that I am scared for the health of my friends and family, of course. This is very much just my personal perspective.

It’s mainly this feeling of uncertainty. Things are changing so fast and the near future cannot be predicted.

I feel this way especially because, as a software developer, my job is among the jobs that are most affected by the AI revolution. And with time, I see more and more parallels between today and the years 2020 to 2022:

#You are glued to the news

It feels like every day something unprecedented happens and changes everything. Some new virus variant / AI model stirs up expectations, another expert weighs in and fuels concerns, or politicians propose new measures to deal with the situation. And I keep refreshing the websites of news outlets. Staying on top of new developments can give you a (false) sense of control in a chaotic situation.

#Everyone knows hyper-specific technical terms

Language model, open-weights and inference are the new virus variant, aerosol and incidence rate. It’s like when everybody becomes an armchair football manager during the World Cup.

#Some jobs are “over”

Some jobs are being written off as doomed. In the pandemic these might have been retail salespeople or anything related to parties and concerts. In the age of AI, it’s software developers, translators, illustrators, designers, musicians, actors, accountants… These speculations trigger a genuine fear of losing status among those affected.

#Social norms constantly need to be renegotiated

“I had Covid one week ago. Is it okay if I come over?” “Do we shake hands to say hello?” During the pandemic we needed to assess what was still okay to do and what was deemed inappropriate over and over again. (I was once scolded by my boss for going for a walk in the forest during the pandemic.)

These days there are countless discussions online and offline about what’s an appropriate way to handle the limitless stream of information that LLMs produce. A sentence that starts with “But ChatGPT said…” can cause a spontaneous discussion of principles at any time. This is exhausting.

#Stuff gets expensive

Although the reasons are completely different this time around, electronics are becoming more and more expensive again. During the pandemic this was mainly caused by disrupted supply chains and lockdowns in China. Today it’s mainly the unprecedented demand created by AI infrastructure projects.

#The youth gets hit hard

The pandemic robbed the youth of education opportunities by forcing them to stay at home while in many countries the education system was not at all prepared for it. Today, it’s just too easy for the youth to rob themselves of education opportunities by using AI for all kinds of school tasks, while teachers are scrambling to find ways to deal with this issue.

#Did it escape from a lab?

…the big difference here being that this time, the scientists are very eager to let you know that their creation has indeed escaped the lab and is very dangerous.

What other parallels do you see between the two?